Diffeomorphic Independent Contrasts for Ancestral Reconstruction of Shapes

Area of Science:

• Evolutionary Biology
• Morphometrics
• Phylogenetics

Background:

• Ancestral state reconstruction is crucial for understanding evolution but current methods oversimplify shape relationships.
• Existing approaches often fail to incorporate landmark correlations within anatomical features.
• Linear assumptions in current models limit their accuracy in capturing complex morphological changes.

Purpose of the Study:

• To introduce a novel method, DICAROS, for accurate ancestral shape reconstruction.
• To combine Large Deformation Diffeomorphic Metric Mapping (LDDMM) with Felsenstein's Independent Contrasts (IC) for improved ancestral state estimation.
• To validate DICAROS against existing methods and apply it to Lepidoptera morphology.

Main Methods:

• Developed Diffeomorphic Independent Contrasts for Ancestral Reconstruction of Shapes (DICAROS).
• Integrated LDDMM for smooth, invertible shape transformations with landmark preservation.
• Utilized Felsenstein's Independent Contrasts (IC) for iterative ancestral shape reconstruction on phylogenetic trees.

Main Results:

• DICAROS demonstrated superior accuracy compared to Ordinary Least Squares, Maximum Likelihood under Brownian Motion, and ridge archetypal penalized likelihood.
• The method successfully reconstructed ancestral shapes and visualized evolutionary trajectories in swallowtail butterflies (Papilionidae).
• DICAROS accurately depicted the evolutionary transition between untailed and tailed Papilionidae species.

Conclusions:

• DICAROS offers a more accurate reconstruction of ancestral shapes, particularly for non-symmetric phylogenetic trees.
• The method provides a robust framework for analyzing complex morphological evolution.
• DICAROS enables visualization of evolutionary trajectories and hypothetical ancestral morphologies.